README

A set of python tools to analyze experimental and simulation data of system showing avalanches, both as a temporal process and in images

How to use it
----------------------------------
The code is still at a preliminar stage and must be used with ipython.
I suggest to use the -qt4 framework (ipython -q4thread), as the plots give less problems than with other threads.

The image files to test the code can be downloaded here:

http://emma.inrim.it/files/testimages.tgz

there are 1500 images; the real first image, with a small avalanche, is at 0663. 
Clearly this can optimized cutting all the first 662 images (see below), 
but for test purpouses it could be nice to have the whole sequence.

So under ipython this is the sequence of commands:

>>> run visualBarkh.py

>>> mainDir = "\yourDirOfData"
>>> firstImage = 0 # Or put any number you like, such as 650
>>> lastImage = None # It takes up to the last image
>>> images = StackImages(mainDir,sigma=2.5,imageFirst=firstImage, imageLas  t=lastImage)

The the code loads and prepares a 3D numpy array of the type images(imageNumber, dimX, dimY).

There are a few methods now that can be used:

>>> images.Array, which contains the whole array
>>> images.shape(), which gives (1499, 500, 500) in this case
>>> images.imDiffShow(940,939), to see the difference between image 940 and 939

>>> images.pixelTimeSequenceShow((320,211)), giving the gray_scale sequence (1500 pts) of pixel = 320, 211

To calculate the avalanches, and get the color picture similar to the korean plots first run:

>>> images.getColorImage()

Here there is the core of the calculation, and the part to be parallelized (see TODO file).

To really see the color image now run, for instance:

>>> images.showColorImage(4, palette='random')

>>> images.showColorImage?
    
    showColorImage([threshold, palette, noSwitchColor, ask])
    Show the calculated color Image of the avalanches.
    Run getColorImage if not done before.
    
    Parameters
    ---------------
    threshold: integer, optional
        Defines if the pixel switches when gray_level_change >= threshold
    palette: string, required, default = 'korean'
        Choose a palette between 'korean', 'randomKorean', and 'random'
        'randomKorean' is a random permutation of the korean palette
        'random' is calculated on the fly, so each call of the method gives different colors
    noSwithColor: string, optional, default = 'black'
        background color for pixels having gray_level_change below the threshold